Hello Oleg,

Thanks a lot for your reply. And you are right, counterpoise correction is the correction term for interaction energy. I read some literature and got clarified about BSSE. But I have a few more questions. What exactly is done in geometry optimization? Wont it disrupt any weak interactions or generate new interaction during the process? Also, if want to know the strength of a particular hydrogen bond in the crystal structure, what should be the most optimum way to do so?

Please let me know if I am thinking in a wrong way.

I agree with Oleg's comments. In general: you always need to do geometry optimization, since you have to make sure that your system in on the potential energy surface minimum. If it isn't, calculating any chemical/physical properties doesn't make any sense.

In your case if it's crystal, you should build a system with PBC and optimize it. Now, this can be quite expensive (computationally) and time-consuming depending on which code you're using. On the other hand you claim that the solution dimer structure is similar and the Hbonds are strong. If this is the case, optimizing a dimer seems a reasonable approximation. Do that, check if the dimer structure is similar to the crystal structure (you can calculate the rmsds of the heavy atoms to get a good measure) and if so, you can proceed with the interaction energy calcualtions using the standard scheme (E(optimized dimer) - E(optimized monomers) + BSSE). Make sure, though, that both the dimer and the monomers are true minima (calculate second derivatives / frequencies).

If the optimized structure of the dimer is different from the structure in the crystal, switch to a PBC code/software.

Hi  :-)

It's necessary to optimize this structure to obtain correct values of angles and distances. The best way to optimize crystal structure is using the function in Gaussian 6-31g* with the polzarization.

So you have to write in the first line in Gaussian input file the following commends:

 hf 6-31g* freq opt

additional function freq allows to obtain correct frequenties of the atomic movement . Remember to read in log file optimized values and GOOD LUCK  because Gaussian is great :-) if you want , i can send you input files. let me know   

Hi Oleg and Bartosz,

Thanks for the suggestions! I read about using PBC and it seems to be out of my reach because of the computational cost involved. The dimer model is working quite fine in my case preserving the strongest interactions (hydrogen bonds) predicted from solution state experimental data as well as those observed in crystal structure. As Oleg mentioned, there could be some different interactions also but I have not analysed those yet.

Hi Agata,

I tried the commands you suggested. In stead of 6-31g*, I used 3-21g.

In some of the references I have seen people using mp2/6-31+g(d,p) for similar calculations. But I used hf/3-21g for initial optimization because using the high level of theory directly was not a good idea. The question I would like to ask you is, for initial optimization should I use the polarization function or the optimization without polarization should also be fine?

Mr. Khare,

If you have performed optimization of the unit cell content of your crystal, than the optimized molecular ensemble becomes unrepresentative to the real crystal structure.

The optimization has yielded to both change of the intermolecular hydrogen bondings and the geometry of the molecules.

So that, the parameters obtained have expressed given state, corresponding presumably to the global minimum, but this state has expressed interactions, different from those ones, observed in the real crystal state, which, in fact, you have determined, experimentally by single crystal X-ray diffraction.

Please find as attachment the molecular structure and optimized ensemble of species using the crystallographic coordinates. As you can see the interaction mode has been changed, which reflects to the obtained vibrational spectra as well. In spite of the fact that they are not scalled, the experimental IR - spectra of both the crystal and it deuterated derivative are given, so that there is possible to obtain the f1 exactly and to determine quantitatively the different frequency positions.

If you would like to receive real estimation of given to crystal parameters, you should keep the structural data as they have been refined experimentally. Furthermore, the optimizatioin cannot express polymorphs as well, which also have yielded different spectroscopic profiles.

It has shown refs in this topic in the attachment, but we have performed in a more recent book correlation between the quantum chemical and experimental parameters for more complex crystal systems as an effort to discuss, exactly applied aspects of the quantum chemistry to problematic joined with the systems, existing and stable under real environmental biogeochemical conditions.

The book is:

B.Ivanova, M. Spiteller, Theoretical design for environmental sorption processes of actinides, LAP Lambert Academic Publishing - ISBN: 978-3-659-63487-1, 2014, pp.1 - 268

You most probably may find in this book something useful, regarding your question as well.

Hello Ms. Ivanova,

Thank you for your detailed reply. It is interesting to see in your attachment how crystal structure coordinates differ from the optimized dimer. In conclusion, I infer from your explanation, that the optimization of crystal structure coordinates should not be done as a dimer but as a lattice using periodic boundary conditions, which is more likely to give a scenario closest to the real crystal structure.

With the optimized structures of the dimer you can evaluate the hydrogen bonds interactions by means AIM analysis with the AIM2000 program. If you have more information about the program write me an e-mail. Best wishes.

Hello Diego,

Thanks for your reply. I do not have AIM2000 program but I have Gaussian09. I tried to run gaussian with AIM keyword and got some output related to AIM analysis. I will go through the output and come back to you if I have any doubts.

Mr. Khare,

In my attachment the first picture exactly represents the disposition of the molecules in the crystal structure (glay) and how the crystallographic packing is changed after optimization (red).